⚙
Build real technology
Design sensors, PCBs, 3D-printed hardware, calibration rigs, embedded systems, camera rigs, and experimental tools that can actually be deployed.
High school + college founding network
Protect the
night we share.
Project Nightfield is a student-led network combining engineering, AI, data science, field research, astronomy, and conservation to understand nighttime environments, build tools that make them measurable, and turn evidence into real local improvement.
Build hardware
Train models
Analyze data
Run field research
Create real-world change
FIELD
ENGINEER
VERIFY
ANALYZE
Why join Nightfield?
Come here to build, experiment, measure, and ship.
Nightfield is for high school and college students who want something more hands-on than another awareness club. You can join as a builder, coder, researcher, data person, field scientist, designer, or organizer and own a real piece of the system.
⌘
Code something useful
Work on web infrastructure, data pipelines, GIS, Nightfield Atlas, visualization, firmware, automation, or open-source research tools.
◈
Train and test AI
Explore machine learning, computer vision, deep learning, TinyML, site-prioritization models, and rigorous evaluation on real environmental problems.
∿
Work with messy real data
Calibration, uncertainty, sensor drift, geospatial data, satellite observations, field measurements, validation, and before-versus-after analysis all matter here.
✦
Get outside the screen
If you prefer fieldwork, run approved Night Audits, study actual sites, meet community stakeholders, and help move a recommendation toward implementation.
↗
Shape something early
Nightfield is still in its founding pilot. Early contributors can help define the hardware, research methods, software, datasets, and culture rather than inheriting a finished system.
You do not need to fit one stereotype.
A CAD student, Python programmer, electronics builder, machine-learning researcher, statistics person, astronomy student, environmental scientist, outgoing organizer, or someone who simply likes experimenting can all contribute to the same mission.
The operating idea
Don't just document a problem. Try to change a place.
Nightfield connects scientific evidence to action. Whether the evidence comes from a field observation, a student-built sensor, a GIS pipeline, or an ML model, the loop stays the same.
01
Measure
Collect structured observations, sensor readings, images, environmental context, or remote data using a documented method.
02
Act
Turn the strongest evidence into a realistic recommendation for the responsible school, campus, park, municipality, or organization.
03
Verify
Return after a change, repeat a comparable measurement, and distinguish real improvement from assumption.
Baseline
→
Intervention
→
Re-measurement
Two major pathways
Go into the field. Or build what the field needs.
Neither path is secondary. Nightfield needs people who can collect credible evidence and people who can invent the tools, software, models, and infrastructure behind that evidence.
Nightfield Field Network
Study schools, parks, campuses, observatories, nature areas, and community spaces. Record conditions, identify something actionable, work with the responsible organization, and return for follow-up evidence.
Nightfield Labs
Build sensors, mechanical systems, electronics, calibration tools, GIS infrastructure, data pipelines, AI, deep-learning models, and open technology that makes environmental fieldwork more scalable.
The guides are starting points, not rigid curricula.
Use them to understand Nightfield's current method, priority project ideas, data-quality expectations, and safety boundaries. Then improve them through real pilot experience.
Nightfield Labs
This is a technology project too. A big one.
Nightfield Labs is the engineering and computational core of the network. We're looking for students who like building prototypes, breaking assumptions, calibrating sensors, writing code, working with imperfect data, training models, and turning experiments into usable tools.
Start here before claiming a Labs project
Nightfield Labs + Engineering Starter Guide
This is a helpful project template, not a rulebook. It contains the current build ideas we're actively interested in, possible MVPs, testing expectations, AI/ML directions, hardware pathways, and ways different teams can connect. Read it first so you can improve an existing direction instead of accidentally duplicating work.
Priority hardware
NightNode Mini
Build a low-cost, open nighttime environmental sensing node for repeatable light, temperature, humidity, and optional privacy-conscious acoustic measurements.
ESP32 • CAD • sensors
NEEDED
Priority research
Calibration Station
Test inter-device variation, linearity, repeatability, temperature drift, enclosure effects, and correction models so cheap sensors can produce more comparable data.
Metrology • statistics
NEEDED
Priority platform
Nightfield Atlas
Build an interactive map combining Nightfield sites, audits, interventions, NightNode data, and properly attributed remote nighttime-light context.
GIS • web • data
NEEDED
Deep learning
NightVision
Develop a computer-vision screening system for standardized, permission-cleared fixture imagery, with site-level testing, uncertainty, and human review built into the workflow.
CV • transfer learning
NEEDED
Machine learning
Site Prioritizer
Use environmental, geospatial, and remote features to help identify where limited field time could generate the most useful new evidence.
ML • GIS • data
RESEARCH
Mechanical / optics
NightShade
Build a safe low-voltage experimental lighting rig with interchangeable 3D-printed shields and quantitatively compare how geometry redirects light.
CAD • optics
BUILD
Embedded AI
Acoustic Node
Explore DSP and TinyML using local spectral features and broad environmental sound categories without making raw conversation storage the default.
DSP • TinyML
EXPERIMENTAL
Electrical
NightNode PCB
Once a prototype is actually validated, translate it into a reproducible PCB with documented sensors, connectors, test points, power, and hardware revisions.
KiCad • EE
LATER STAGE
Imaging hardware
NightVision Rig
Design a repeatable phone or camera mount, angle procedure, and metadata workflow that reduces uncontrolled variation in computer-vision datasets.
CAD • imaging
DATA QUALITY
A Nightfield technical project should solve something real.
A project counts when it improves measurement quality, reduces field effort, helps choose better sites, supports an intervention, improves data reliability, or makes an environmental result easier to verify. "Used AI" by itself is not the goal.
Nightfield Field Network
One place. A repeatable investigation.
The Night Audit is designed to be approachable enough for a new team while still producing structured evidence that can connect to Nightfield Labs.
Field starter template
Night Audit Kit v0.2
Use this guide before your first audit. It is intentionally a starter template: keep the core method consistent, then improve the workflow after real pilot feedback. It also explains how field teams can contribute data to Atlas, NightNode, NightVision, and future ML work.
✦
Sky visibility
Use an established citizen-science observing method instead of inventing an arbitrary Nightfield darkness score.
◉
Outdoor lighting
Document targeting, spill, glare, controls, operation, and other potentially correctable conditions.
⌁
Site environment
Record habitat, land use, astronomy context, local constraints, and relevant environmental observations.
≋
Optional soundscape
Document broad nighttime sound sources and connect future quantitative work to validated or calibrated setups.
One network
The interesting part is when the disciplines collide.
A mechanical student, embedded developer, ML researcher, GIS contributor, field team, and outreach lead can all touch the same project before it becomes a verified result.
From signal to change
This is the long-term Nightfield pipeline we're trying to build.
01
Prioritize site
02
Audit / deploy
03
Analyze data
04
Pursue action
05
Measure again
06
Verify result
Open impact
Measure outcomes, not membership.
Nightfield's dashboard is a working template during the founding pilot. The point is to eventually report what actually happened: audits, validated builds, interventions, follow-ups, and verified improvements.
Founding pilot dashboard
The spreadsheet is a starting template. Metrics should populate only as reviewed work is completed.
Audits
completed investigations
Builds
validated technical work
Actions
interventions pursued
Verified
follow-up improvements
Founding team
Leadership means owning an output.
Nightfield is deliberately avoiding a giant hierarchy of decorative titles. The founding team should stay small, technical where appropriate, and accountable for actually shipping work.
Strategy
Project Lead
Coordinates direction, execution, technical development, field science, and partnerships.
Science
Science & Field Methods
Maintains audit methodology, data-quality expectations, and follow-up standards.
Engineering
Nightfield Labs
Coordinates NightNode, CAD, electronics, calibration, and engineering projects.
AI / research
AI & ML
Coordinates NightVision, site-prioritization, model evaluation, and responsible ML practice.
Data
Atlas & Data
Builds GIS, data pipelines, analytics, database structure, and Nightfield Atlas.
Infrastructure
Web & Infrastructure
Maintains the site, repositories, forms, automation, and digital systems.
Growth
Network & Outreach
Finds strong contributors and helps them become active teams instead of passive members.
Relationships
Partnerships
Builds legitimate relationships with educators, astronomy groups, researchers, and conservation organizations.
Communication
Media & Storytelling
Turns completed experiments, builds, audits, and interventions into clear public stories.
Scientific backbone
Use existing science where it works. Build where there's a gap.
Globe at Night
Field teams can use the official Globe at Night workflow for citizen-science sky observations instead of Nightfield inventing its own unsupported scale.
Third-party program.
Reference does not imply affiliation or endorsement.
Responsible lighting
Nightfield field screening can reference the Five Principles for Responsible Outdoor Lighting: Useful, Targeted, Low Level, Controlled, and Warm-Colored.
Framework attributed to DarkSky International
and the Illuminating Engineering Society.
Nightfield remains independent.
Open Earth data
Nightfield Labs may use appropriately licensed and attributed Earth-observation and mapping resources for remote environmental context.
Datasets and map sources must be attributed under their applicable terms.
External resources do not imply sponsorship.
Privacy, safety & transparency
Serious project. Clear boundaries.
Project Nightfield Network is currently an independent, student-led, noncommercial initiative.
Privacy & personal information
Minimal collection. Nightfield should collect only information reasonably needed to coordinate participants, review contributions, operate teams, and measure project impact.
Participants should not submit home addresses, government identification, financial information, passwords, private schedules, or other unnecessary sensitive information.
Public impact reporting should use aggregate, site-level, or non-identifying information whenever practical. Private participant contact information should not appear on the public impact dashboard.
Nightfield does not sell participant personal information.
Questions concerning information controlled directly by Nightfield may be sent to nightfieldnetwork@gmail.com .
Student participation
Participants under 18 should follow applicable parent or guardian, school, club, property, and site requirements.
Do not publicly post another participant's private phone number, home address, private email, schedule, or unnecessary identifying information.
Identifiable photographs, videos, or quotations involving minors should not be publicly published by Nightfield without appropriate permission.
Field & engineering safety
Night Audits are observational activities and should occur only in safely accessible, permitted locations.
Do not trespass, enter restricted areas, stand in roadways, climb lighting structures, open installed electrical equipment, interfere with wildlife, or physically modify public lighting.
Nightfield Labs student hardware projects should remain within appropriate low-voltage electronics, mechanical prototyping, software, data science, and supervised research activities.
Installed mains-powered lighting should be handled only by appropriately responsible or qualified personnel.
AI / ML transparency
Nightfield AI systems are experimental research and screening tools.
NightVision and related models should not be represented as professional engineering inspectors, regulatory tools, or replacements for human review.
Model evaluation should use appropriately separated data where relevant and should document limitations, uncertainty, and failure cases.
A flashy training-set accuracy number is not evidence that a model works in new communities, new cameras, or new environments.
External organizations, datasets & attribution
References to external organizations, programs, datasets, methods, or frameworks do not imply sponsorship, funding, certification, endorsement, or formal partnership.
Third-party names, logos, trademarks, datasets, photographs, publications, and other materials remain subject to the rights and licenses of their respective owners.
Nightfield should not use another organization's logo as Nightfield branding or claim a partnership until that relationship actually exists.
NASA Earth-science datasets used in Nightfield work should be cited at the dataset/product level where appropriate.
If OpenStreetMap data is used in Nightfield Atlas, the required contributor attribution and applicable Open Database License notice must be displayed.
The Five Principles for Responsible Outdoor Lighting should remain properly attributed to DarkSky International and the Illuminating Engineering Society.
Scientific limitations
Student observations, experimental models, and prototype sensors may contain measurement uncertainty, device variability, observer variation, environmental effects, calibration error, or other limitations.
Nightfield observations are not professional electrical inspections, engineering certifications, regulatory determinations, or guarantees of site safety.
Published work should distinguish between raw measurement, analysis, prediction, recommendation, implemented action, and verified result.
Participation terms
Project Nightfield Network is currently an independent, student-led, noncommercial initiative. Participation and leadership roles are currently unpaid.
Participation does not guarantee employment, financial compensation, academic credit, college-admission benefit, research authorship, or leadership status.
Responsibility should follow real contribution, reliability, judgment, and continued involvement.
Founding network now forming
Don't just join a club. Help build the system.
Pick a site. Design a sensor. Build a map. Train a model. Run an experiment. Analyze a dataset. Talk to a community. Fix a broken method. Start with one useful contribution.